Multi-Source Data for Short-Term Influenza Forecasting in Shenzhen, China (2023–2025)

A 16.8 KB document by Xing Li, published on figshare in April 2026, presents a generalized additive model for influenza forecasting. The model integrates surveillance, mobility, meteorological, and search index data from 2023 to 2025 for Shenzhen, China. It achieved an R² of 0.85 for 1-week ahead forecasts and was validated against a SARIMAX model.

Use Cases

• Benchmarking time-series forecasting models based on the described multi-source data integration approach.
• Studying the impact of cross-boundary mobility on disease spread based on the Hong Kong-Shenzhen mobility metric.
• Developing interpretable public health models based on the generalized additive model (GAM) framework described.
• Analyzing the relationship between search engine queries and influenza activity based on the Baidu Search Index data.

Strengths

• Model performance is quantified with specific metrics, including an R² of 0.85 for 1-week forecasts.
• Data integrates multiple sources, including surveillance, mobility, weather, and search trends.
• The model was validated against a SARIMAX benchmark, demonstrating comparative performance.
• The dataset is openly licensed under CC-BY-4.0.

Limitations

• The dataset is a 16.8 KB document; the underlying tabular data is not directly provided.
• Column-level documentation is absent; field semantics must be inferred from the paper.
• Row count is unknown, which may limit suitability assessment for direct ML training.

Provenance

Source

figshare

Collection Method

Integrates surveillance and auxiliary data, likely from public health and commercial sources.

Time Range

2023–2025

Freshness

Last updated 2026-04-30 13:14:56; freshness should be verified.

Geography

Shenzhen, China, with cross-boundary data from Hong Kong.